ElasticSearch at Ticketbis: A real story in +50 languages
summary
Ticketbis describes their heavy use of ElasticSearch (about 100M indexed documents plus ~2 billion log documents) and explains how they performed capacity planning for clusters—deciding number of nodes, indexes, shards, and replicas. They discuss the added complexity of indexing content in over 50 languages and the importance of choosing the correct multilingual approach, and they also show how they store ticket price variations and use that data to analyze seller behavior trends.
AI-written summary · about
chapters (12)
- 0:00Introduction
- 1:35ElasticSearch at Ticketbis
- 3:42Capacity planning
- 8:16Growth analysis and shards
- 14:22Real search use case
- 15:48Internationalization challenges
- 18:36Multilingual data models
- 21:03ElasticSearch type problems
- 23:12Language-specific indices
- 26:46Final decision and scalability
- 28:09Conclusions and recommendations
- 29:13Q&A
related talks
35mElasticsearch - Beyond Full-text Search
37mHow we solved for Aggregation, Search and Time Series Big Data usecases using Elasticsearch Database
16mWhy we built ElasticSearch
54m7 Petabytes of Logs in Elastic. How We Did It
31mUsing Elasticsearch with Rails Apps
29mHow (not) to Scale Elasticsearch for Data Analytics!
24mOperating at scale with Elastic Search in production
44mElasticsearch at Scale of Billions of Documents